With slower declines in incidence and continuing high mortality rates associated with colorectal cancer (CRC) among Latinos, developing and testing culturally-specific programs designed to improve prevention behaviors in this population becomes increasingly important. Two known CRC risk behaviors prevalent among Latinos are low levels of physical activity and low consumption of vegetables. Specific Aim 1. To evaluate effects of a narrative theory-based intervention compared to a standard intervention on subsequent CRC prevention behaviors and screening in a low-income Latino population. A two-group randomized, controlled trial is proposed to examine effects of a novel, culturally aligned Latino storytelling (ST) intervention versus a standard didactic learning (DL) intervention communicating cancer risk, primary prevention and screening information in the community setting, (900 participants, ~450 in 80 groups per arm of study). Both intervention groups will receive cancer risk and prevention information, delivered by Lay Health Educators (LHEs), styled according to arm of study and subsequent biweekly sessions over 12 weeks. Outcomes will be assessed post-intervention and 6 months later to examine maintenance. H1: Latinos will show greater increases in primary CRC prevention behaviors based on their exposure to ST intervention relative to those who are exposed to DL intervention. Primary endpoints are (a) minutes of physical activity per week, and (b) daily servings of vegetables, both measured with highly reliable, validated instruments. Secondary endpoints include additional CRC risk-related behaviors (calcium intake, red meat consumption, obesity, alcohol, tobacco use) and screening (FOBT and endoscopy). Characteristics of LHEs and core informational content will be consistent across both arms of intervention;only the style and context of delivery will be radically different. Specific Aim 2. To test specific theory-based models that compares the role of each intervention on the physical activity and dietary outcomes. Influences of theoretical constructs associated with this method will be tested in two primary models. First we examine the difference in effects of interventions on behavior predictive factors suggested by Theory of Planned Behavior. Then elements of a proposed explanatory model of the ST intervention, including transportation, identification, self-identity and reciprocal support factors, will be tested for mediation effects on outcomes.